Huang, T., Zanocco, C., Wang, Z. et al. Built environment disparities are amplified during extreme weather recovery. Nature
Objective:
- Produce one of the largest and most granular disaster recovery datasets to date by using multimodal machine learning to show building-level trajectories of post-disaster recovery with street view image
Case:
- US census tract
Methodology:
- ML
Data Source
- Google street view
- FEMA weather events
Findings:
- Buildings in lower-income communities have higher rates of becoming empty lots and lower improvment rates
